Endocrine-Related Cancer (2006) 13 95–111
A molecular mimic demonstrates that phosphorylated human prolactin is a potent anti-angiogenic hormone Eric Ueda1,2, Ugur Ozerdem3, Yen-Hao Chen1, Min Yao4, Kuang Tzu Huang1, Huiqin Sun3, Manuela Martins-Green4, Paolo Bartolini 2 and Ameae M Walker1 1
Division of Biomedical Sciences, University of California, Riverside, California 92521, USA IPEN-CNEN-Biotechnology Department, Universidade de Sao Paulo, Sao Paulo, CEP 05508–900 Brazil 3 La Jolla Institute for Molecular Medicine, San Diego, California 92121, USA 4 Cell Biology and Neuroscience, University of California, Riverside, California 92521, USA 2
(Requests for offprints should be addressed to A M Walker; Email:
[email protected])
Abstract S179D prolactin (PRL) is an experimentally useful mimic of naturally phosphorylated human prolactin. S179D PRL, but not unmodified PRL, was found to be anti-angiogenic in both the chorioallantoic membrane and corneal assays. Further investigation using human endothelial in vitro models showed reduced cell number, reduced tubule formation in Matrigel, and reduced migration and invasion, as a function of treatment with S179D PRL. Analysis of growth factors in human endothelial cells in response to S179D PRL showed: a decreased expression or release of endogenous PRL, heme-oxygenase-1, basic fibroblast growth factor (bFGF), angiogenin, epidermal growth factor and vascular endothelial growth factor; and an increased expression of inhibitors of matrix metalloproteases. S179D PRL also blocked signaling from bFGF in these cells. We conclude that this molecular mimic of a pituitary hormone is a potent anti-angiogenic protein, partly as a result of its ability to reduce utilization of several well-established endothelial autocrine growth loops, partly by its ability to block signaling from bFGF and partly because of its ability to decrease endothelial migration. These findings suggest that circulating levels of phosphorylated PRL may influence the progression of cancer and, furthermore, that S179D PRL may be a useful anti-angiogenic therapeutic. Endocrine-Related Cancer (2006) 13 95–111
Introduction Many tissues express prolactin (PRL) receptors (Costlow & McGuire 1977, Meister et al. 1992, Royster et al. 1995), indicating the likelihood of a response to PRL, the primary source of which is the pituitary. In addition, a substantial proportion of these tissues express PRL (reviewed in Ben-Jonathan et al. 1996) and for some, such as the mammary gland and prostate, PRL functions as an autocrine growth factor (Ben-Jonathan et al. 1996, Clevenger & Plank 1997, Schroeder et al. 2003). An autocrine PRL growth loop is present in cells representative of both steroiddependent and -independent breast and prostate cancer (Clevenger & Plank 1997; reviewed in Ben-Jonathan et al. 2002), and is a potential target for therapeutics.
Since agents that regulate pituitary PRL synthesis and release are ineffective in extrapituitary tissues (BenJonathan et al. 1996), several laboratories have developed PRL receptor antagonists (Fuh & Wells 1995, Lochan et al. 1995, Chen et al. 1998, 2002, Llovera et al. 2000, Xu et al. 2001, Beck et al. 2003). S179D PRL, a mimic of phosphorylated human PRL (Wang et al. 1996, Chen et al. 1998, Tuazon et al. 2002), was originally developed as a PRL receptor antagonist since we had shown that naturally phosphorylated PRL had this activity (Krown et al. 1992, Wang & Walker 1993). It was anticipated therefore that S179D PRL would impair the growth of tissues utilizing PRL as a growth factor by blocking the effects of both pituitary and autocrine PRL. Indeed, this molecule has been shown to impair the development of a number of fetal tissues (Coss et al.
Endocrine-Related Cancer (2006) 13 95–111 1351-0088/06/013–95 g 2006 Society for Endocrinology Printed in Great Britain
DOI:10.1677/erc.1.01076 Online version via http://www.endocrinology-journals.org
E Ueda et al.: Phosphorylated prolactin is anti-angiogenic 2000, Yang et al. 2001, 2002), to inhibit the growth of mammary glands during pregnancy (Kuo et al. 2002, Naylor et al. 2005), to inhibit proliferation of human breast cancer cells in response to unmodified PRL (U-PRL) in vitro (Schroeder et al. 2003) and to inhibit the growth of human prostate cancer cells both in vitro and when grown as tumors in nude mice (Xu et al. 2001). Further analysis of the effects of S179D PRL, however, made two points clear: first, that S179D PRL was not a pure antagonist (Bernichtein et al. 2001, Wu et al. 2003, Xu et al. 2003, Tan et al. 2005); and second, that it was more effective in vivo than in vitro (best illustrated in Xu et al. 2001). With regard to the first issue, we have demonstrated that in addition to its growth-inhibiting properties, S179D PRL promotes the expression of tissue-specific genes in both the mammary gland (Kuo et al. 2002, Wu et al. 2003) and the prostate (Xu et al. 2003). These dual antagonist/agonist activities are brought about by the inhibition of signaling to cell proliferation (Coss et al. 1999, Schroeder et al. 2003, Wu et al. 2003) and by production of alternative signals which result in the expression of tissue-specific genes (Wu et al. 2003) and cell cycle regulatory proteins (Wu et al. 2005). With regard to the greater efficacy in vivo (Coss et al. 2000, Xu et al. 2001, Kuo et al. 2002, Naylor et al. 2005), one potential explanation is that S179D PRL acts on a variety of cells and not just the mammary and prostate epithelial cells initially studied. Among these potential targets, cells of the developing vasculature are good candidates. In addition to being a requirement for normal development, angiogenesis is also an essential element of many pathological processes, including tumor growth and metastasis, diabetic retinopathy and neurofibromatosis type 1 (reviewed in Folkman 1995, Ozerdem 2005). The development of anti-angiogenic therapies for treatment of these pathologies has therefore become an increasingly important goal of biomedical research. Endothelial cells have been shown to express PRL receptors (Corbacho et al. 2000, Merkle et al. 2000) as well as a unique receptor that binds a cleaved form of PRL (Clapp & Weiner 1992). We therefore undertook the current study to determine whether part of the in vivo anti-growth efficacy of S179D PRL was the result of anti-angiogenic activity.
Materials and methods Recombinant prolactins Recombinant U-PRL and S179D PRL were produced as previously described (Chen et al. 1998). Endotoxin
96
levels in the preparations were detected and measured with the Limulus amoebocyte lysate assay (E-Toxate kit; Sigma).
Angiogenesis assays in the chicken chorioallantoic membrane (CAM) These assays were carried out using a modification of the method previously described (Martins-Green & Kelly 1998). Briefly, 2 ml albumin were withdrawn from 4-day fertilized eggs to drop the CAM away from the shell. Windows in the shell were made 3 h later and the eggs were incubated until day 9. Pellets were prepared by mixing 1 mg U-PRL or S179D PRL, or 500 ng basic fibroblast growth factor (bFGF), with 1% methylcellulose (1 : 1) in a total volume of 40 ml per pellet. Then, 40 ml drops were placed on parafilm and allowed to dry in a sterile hood for 2 h. This procedure yielded small disc-shaped pellets, 5 mm in diameter and 0.1 mm thick. These were then placed between two large blood vessels of the 9-day-old CAMs; 5 days later, the pellets and the surrounding areas were removed, fixed and the CAMs examined for the degree of angiogenesis. Angiogenesis was quantified by estimation of the avascular area. Briefly, using enlarged photographs of the pellet circles, the areas that did not contain the normal frequency of smaller vessels present in the positive controls were cut out and weighed. Weights were converted to area knowing the magnification of the images. The positive controls with normal spacing of the small vessels were considered to have no avascular area. Negative controls consisted of pellets with water and methylcellulose only, and pellets containing both S179D PRL and a 3-fold excess of rabbit anti-human PRL (provided by A Parlow, Harbor General, Torrance, CA, USA). The bFGF pellet served as a positive control. The experiment was carried out five times with ten replicates for each treatment.
Corneal angiogenesis assay The surgical procedure for inducing corneal angiogenesis in the mouse (Kenyon et al. 1996) was modified to incorporate two pellets in the corneal pocket instead of just a single pellet (Ozerdem & Stallcup 2004). Slow-release polyhydroxyethyl methacrylate (hydron) (Hydro Med Sciences, Cranbury, NJ, USA) pellets (0.4r0.4r0.2 mm) were formulated to contain 45 mg sucrose aluminum sulfate (sucralfate) (Sigma) plus one of three experimental additives: 90 ng recombinant bFGF (Life Technologies, Carlsbad, CA, USA), 52.5 ng S179D PRL or 52.5 ng U-PRL. Ten mice (6 weeks old; C57BL/6) were anesthetized www.endocrinology-journals.org
Endocrine-Related Cancer (2006) 13 95–111 with Avertin (0.015–0.017 ml/g body weight) and two pellets were implanted into the corneal stroma at a distance of 0.7 mm from the corneo-scleral limbus. Ten eyes received pairs of pellets containing bFGF and S179D PRL. Fellow eyes of each mouse received pairs of pellets containing bFGF and U-PRL. Over an 8-day period after surgery, the mice were examined to evaluate the progress of corneal angiogenesis in the operated eyes. On day 8, angiogenesis was quantified by determining the area of vascularization, as described previously (Kenyon et al. 1996, 1997).
Cell culture Primary human umbilical vein endothelial cells (HUVEc) and primary human microvascular endothelial cells (HMVEc) were obtained from Clonetics (San Diego, CA, USA). The cells were usually used up to passage 5, but some repeat experiments used cells up to passage 10. HUVEc were grown on collagencoated plates in medium M199 supplemented with 10 mM Hepes, 2.5 mg/ml thymidine, 140 USP units/ml heparin, 5 ng/ml human bFGF (Sigma), 20% fetal bovine serum (FBS; Gibco) as recommended by the supplier. HMVEc were grown on collagen-coated plates in endothelial growth medium with growth factors (EGM-2MV medium) (Clonetics).
Cell proliferation assay HUVEc were seeded in 24-well plates at a density of 3r104 cells/well in 1 ml growth medium. After 24 h, cells were incubated with either U-PRL or S179D PRL in human endothelial serum-free medium containing bFGF, as before, and epidermal growth factor (EGF) at 10 ng/ml (Sigma) (SFM). Cells were further incubated for 72 h at 37 C. Cell number was determined using a colorimetric assay (Cell Titer; Promega) with the stringent modifications described previously (Huang et al. 2004).
Endothelial tubule formation Matrigel with reduced growth factors (BD Biosciences, Palo Alto, CA, USA), diluted 1 : 1 with SFM was added (350 ml) to each well of a 24-well plate and allowed to polymerize fully at 37 C for 30 min. A suspension of 3r104 HUVEc/well in 400 ml SFM with reduced bFGF (2.5 ng/ml) supplemented with 5% FBS was transferred to each well. The cells were then treated with 1 mg/ml of either of the two forms of PRL. Cells were incubated for 24 h at 37 C in a humidified 5% CO2 incubator and the tubule-like structures were observed and quantified in six random microscopic www.endocrinology-journals.org
fields towards the center of the well (at r20 magnification).
DNA content analysis HUVEc were cultured for 3 days in SFM with reduced bFGF (2.5 ng/ml) containing 5% FBS and either UPRL or S179D PRL (1 mg/ml). Cells were harvested by trypsinization, washed with cold Dulbecco’s PBS (DPBS) and fixed with 75% ethanol in DPBS at 4 C for 30 min. Cell pellets were resuspended in 0.1% Triton X-100 (Sigma) in DPBS containing 200 mg/ml RNase (Sigma) and 10 mg/ml propidium iodide (Sigma), and incubated at room temperature for 30 min. The fluorescence of individual cells was measured with a FACScan cytofluorometer equipped with CellQuest software (Becton Dickinson, Franklin Lakes, NJ, USA).
Western blotting and immunoprecipitation To detect PRL receptor (PRLR) isoforms, equal quantities of lysate (1 mg protein) from treated and untreated HUVEc were immunoprecipitated with 1 mg mouse anti-PRLR which recognizes the receptor extracellular domain (Zymed, San Francisco, CA, USA; catalog number, 35–9200); the immunocomplexes were captured with 100 ml protein A-agarose bead slurry (Upstate Biotechnology, Inc., Lake Placid, NY, USA) and thoroughly washed. The beads were then resuspended in 50 ml SDS-loading buffer and, after boiling for 5 min, 20 ml of the supernatant were subjected to SDS-PAGE and subsequent transfer to a polyvinylidene fluoride (PVDF) membrane. AntiPRLR was used at 1 : 1000 dilution. Antigen–antibody interactions were detected using horseradish peroxidase-coupled secondary antibodies and enhanced chemiluminescence (ECL; Amersham). To detect activation of extracellular-regulated kinase (ERK) 1 and 2, cells were cultured in the absence of bFGF for 16 h and then treated with bFGF (25 ng/ml) or bFGF and S179D PRL (1 mg/ml) for the times indicated. Cell lysate proteins were resolved on the gel and antiphosphoERK was used for Western blots (dilution 1 : 1000; Santa Cruz Biotechnology, Inc. Santa Cruz, CA, USA). The blots were stripped and re-probed with anti-total ERK (dilution 1 : 1000; Santa Cruz Biotechnology) to control for loading.
Cloning ring migration assay A cloning ring, used to create a defined initial area of cells, was placed in the middle of a 35 mm Petri dish and a suspension of 104 HMVEc in 100 ml EGM-2MV
97
E Ueda et al.: Phosphorylated prolactin is anti-angiogenic was poured inside the ring and incubated at 37 C for 40 min for cell attachment. After this period, the cloning ring was carefully removed and the edges of the circle formed, were marked. One milliliter of medium (plus the various treatments) was then added. The cells were treated for 3 days and cell migration was assessed on the third day.
Chemotaxis assay HMVEc (106 cells) in 100 ml EGM-2MV medium were plated on the underside of transwell membranes (6.5 mm Transwell filter inserts; membrane pore size, 8 mm; Corning Costar Corporation, Corning, NY, USA) by inverting the insert. The cells were allowed to adhere for 30 min, and then the wells were turned right side up and medium was added to both chambers: 100 ml in the top chamber with 5 ng/ml bFGF alone, or bFGF plus U-PRL or S179D PRL (1 mg/ml); 1 ml in the bottom chamber with EGM2MV only. The cells were allowed to migrate towards the bFGF for 4 h. In order to evaluate migration, the cells on the undersides of the chambers were removed by wiping them away with a cotton swab, and the membranes were then fixed and stained with 2% toluidine blue in 4% formaldehyde. Migrated cells on top of the filter were photographed and counted in ten fields per filter at r10 magnification to obtain the average relative number of migrating cells per treatment (Li et al. 2004).
Establishment of rat aorta culture The proximal aorta was obtained from adult Sprague– Dawley rats as described previously by Wang et al. (2004), and 1 mm rings were sectioned. Twenty-fourwell plates were pre-coated with 200 ml Matrigel (BD Biosciences) diluted 1 : 1 with EGM-2MV. An aortic ring was placed on the top of this layer and covered with additional Matrigel (200 ml). After Matrigel polymerization, 500 ml EGM-2MV containing 2% FBS (medium for preferential growth of endothelial cells according to Wang et al. (2004)) were added to each well and the rings were incubated for 24 h. After this incubation, fresh medium was added with either U-PRL or S179D PRL (1 or 2 mg/ml), and incubation continued for 8–10 days.
bromide) staining. Quantitative assessment used an MTS (3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2H-tetrazolium) assay to measure the relative number of viable cells growing from the aorta rings. After removal of the ring of aorta, 300 ml freshly prepared solution containing 100 mg MTS and 1 mg PMS (phenazine methosulfate) in DPBS was added to each well and incubated for 24 h at 37 C. The medium used for this study preferentially selects for endothelial out-growth, but cell number will include some fibroblasts and smooth muscle cells (Hata-Sugi et al. 2002).
Gene expression by semi-quantitative and real-time RT-PCR Total RNA from non-confluent HUVEc (2r106 cells/ 100 mm dish) was isolated with Trizol (Gibco/Invitrogen). The primers used for RT-PCR are given in Table 1. Expression was normalized to GAPDH mRNA for semi-quantitative RT-PCR, and to b-actin for real-time RT-PCR. Semi-quantitative RT-PCR
Five micrograms of total RNA from each sample were used for cDNA synthesis using Super Script First Strand Synthesis (Invitrogen). A total volume of 2 ml cDNA from the RT reaction was used with the PCR Master Mix (Invitrogen). The amplification program consisted of 95 C for 2 min, followed by 30 cycles consisting of a denaturation step (95 C for 30 s), an annealing step (55 or 56 C) and an extension step (72 C for 30 s). The amplified products were detected by 1% agarose/Tris-borate EDTA (TBE) (0.089 M Tris and borate, 0.002 M Na-EDTA) gel electrophoresis and ethidium bromide staining. Quantitative real-time RT-PCR
Twenty five microliter reactions were prepared in 96-well optical reaction plates (ABI Prism; Applied Biosystems) using SYBR green master mix (Applied Biosystems, Foster City, CA, USA). Using the ABI Prism 7700 (Applied Biosystems), samples were heated to 50 C for 2 min, then 95 C for 10 min, before 40 cycles at 95 C for 15 s and then 55 or 56 C for 1 min. Real-time data were analyzed using the comparative CT method (Bustin 2000).
Qualitative and quantitative measurement of cell out-growth in the aorta ring assay
Protein array
Qualitative assessment was performed as described previously (Wang et al. 2004) with MTT (3-(4,5dimethylthiazol-2-yl)-2,5-diphenyl-2-H-tetrazolium
In order to detect the influence of S179D PRL on the release of angiogenesis-related factors, an angiogenesis antibody array was used (Chemicon International,
98
www.endocrinology-journals.org
Endocrine-Related Cancer (2006) 13 95–111 Table 1 Oligonucleotide sequences employed Primer name
5¢-Oligonucleotide sequence-3¢
PRLR-LF-reverse PRLR-LF-forward PRLR-deletedLF-forward PRLR-deletedLF-reverse PRLR-SF1A-reverse PRLR-SF1A-forward PRLR-SF1B-reverse PRLR-SF1B-forward PRL-EXON2 PRL-EXON5 BETA ACTIN-forward BETA ACTIN-reverse CTGF-forward CTGF-reverse TIMP-1 forward TIMP-1 reverse TIMP-2 forward TIMP-2 reverse IL-8 forward IL-8 reverse VEGF-forward VEGF-reverse HO-1-forward HO-1-reverse TSP-1-forward TSP-1-reverse ANG-2-forward ANG-2-reverse ANG-1 forward ANG-1reverse FGF-forward FGF-reverse ANGI0GENIN-forward ANGI0GENIN-reverse FGFR-reverse FGFR-forward VEGFR-1-reverse VEGFR-1-forward VEGFR-2-reverse VEGFR-2-forward GAPDH-reverse GAPDH-forward
GATTTGATGCTCATCTGTTGGA TCCAGGTATGTGGGTTTCAT ATCATGATGGTCAATGCCACTA TGGGGTTCCTCACACTTTTC GATAGTGAGGACCAGCATCTAATG TGGACTGTGGTCAATGTTGC CATGAATGATACAACCGTGTGG CAACATCAAGGGGTCACCTC GCAGTTGTTGTTGTGGATGATT GATGCCAGGTGACCCTTCGAGA AAAGACCTGTACGCCAACAC GTCATACTCCTGCTTGCTGAT GAGGAAAACATTAAGAACGGCAAA CGGCACAGGTCTTGATGA CTGCGGATACTTCCACAGGTC GCAAGAGTCCATCCTGCAGTT TTGAGAGTGGACCACACTGCGC CTGGCAACCCTACAACAGACCC ATAAGCAGGCCTCCAACGC GAGCTGGACCAGTCGAAACC GCCTTGCTGCTCTACCTCCA CAAGGCCCACAGGGATTTT CAGGCAGAGAATGCTGAGTTC GATGTTGAGCAGGAACGCAGT CCCTTCAAAACAAATAGGAGTTCA ATCCTGTGATTCCAAATGCCAG TGGGATTTGGTAACCCTTCA GTAAGCCTCATTCCCTTCCC GCAACTGGAGCTGATGGACACA CATCTGCACAGTCTCTAAATGGT TGTGCTAACCGTTACCTGGCT CAGTGCCACATACCAACTG TGGGCGTTTTGTTGTTGGTCTTC CGTTTCTGAACCCCGCTGTGG GCCAGCAGTCCCGCATCATCAT GACGCAACAGAGAAAGACTTGT GATGTAGTCTTTACCATCCTG CAAGTGGCCAGAGGCATGGAGTT TGCCAGCAGTCCAGCATGGTCTG GAGGGCCTCTCATGGTGATTGT GGCATGGACTGTGGTCATGAG TGCACCACCAACTGCTTAGC
Inc.Temecula, CA, USA; CHEMIARRAY human angiogenesis antibody array). For this assay, a control consisting of HUVEc cultured in SFM plus bFGF (2.5 ng/ml) and EGF (1 ng/ml), and a treatment group with the same medium plus S179D PRL (1 mg/ml), was used. The conditioned medium was then retrieved after a 24-h incubation. Serum-free medium was chosen to reduce the background quantities of angio-active compounds. Normalization from array to array www.endocrinology-journals.org
utilized the positive control included by the manufacturer for this purpose.
Statistical analysis All numerical data are presented as the meantS.E. Except where noted, all experiments used at least triplicates and were conducted a minimum of three times. Statistical significance was calculated using Student’s t test to compare individual means with or without Bonferroni corrections, as appropriate. A P value of
